View Full Version : Noisy motors + rough cuts

02-17-2003, 07:43 PM
I have recently noticed rough cuts on large circles (11 inch +) in the areas of about 45, 135, 225 and 315 degrees.

Today I happened to make some 'air cuts' without the router running. Both X and Y motors are very rough sounding at these places when both are stepping small amounts. Watching the bit dance confirms that this is the reason for the rough cutting I'm seeing.

Any clues?

Leon Romatowski
02-17-2003, 09:18 PM

-Check if the set screws on your X/Y pinion gears may be loose.

-Could be a rigidity problem on Z, router mount, guide rollers.


02-18-2003, 12:31 PM
Chris look on your original drawing and see if there is some zig zag line at those points You have to zoom way in. sounds simple but this can happen mostly when you do a trace on a scanned image but also sometimes when you draw on the computer due to noise. zoom waaaaaaayyyy in David in Wyoming.Ps if you choose to look at points on the drawing there should only be two or at the most three on a circle.If more then that is the proublem.and you need to clean them up.

02-18-2003, 02:47 PM

Set screws are tight.

Z axis is very rigid & tight.


Problem happens on a single CG or CC command. No extra points.

02-18-2003, 05:16 PM
I have noticed the exact same problem when cutting plywood, for some reason it doesnt happen in MDF?
My 96's Z is solid and I also use the CC command. Not that I cut many circles but it would
be interesting to find out the cause.


02-19-2003, 03:01 AM
The reason for the problem occuring at 45,135,225 and 315 degrees is because the x and y motors are stepping at exactly the same speeds at these points. The "resonance" that they are creating with each other is giving the added roughness. If this resonance is at the same frequency as flexibility of the SB frame, then the situation is much worse. The only things that you can try to do is either to stiffen the frames, or to change the cutting speed to move away from frame resonance.

02-19-2003, 04:17 PM
Why would the speed of the X and Y motors at only those points be any different than the speed at any other point when cutting circles?

I find it hard to believe any "resonance" created would show up as rough cuts in anything as large as 11" circles. What this implies is that the design and construction of the Shopbot cannot handle cutting an 11" circle smoothly. I disagree.

My guess is maybe his ramp settings need to be adjusted. On my PRT things seem to work best with the ramp settings around 60% of the cut speeds. So if I have a cutting speed of 1" per second, my ramps would be set at .60

What are your cut speeds and ramp speeds, Chris, and Wingnut ?

02-20-2003, 03:38 AM
Mayo, if you cut a horizontal line, only the x-motors work and the y-motor stands still. But, if you cut a 45 degree line, both x znd y motors must run at the same speed.

Would you agree to the following if you cut a full circle starting at the 9 o'clock position going clockwise?:


At position 1, the y-motor has maximum speed while the x-motor is just starting to move with near zero speed. But the y-motor must slow down gradually so that it can change direction at position 3. Simultaneously, the x-motor must increase speed so that it can do all the work at position 3. And therefore, at position 2, the two motors have equal speed. This little "dance" (with audible soundtrack) between the motors continues right around the circle and there are 3 more occasions (4,6,8) when then have the same speed.

Another way of looking at this roughness, is to realise that a 45 degree cut requires a 41.4% bigger step than either a pure horizontal, or a pure vertical move. (The diagonal of a square is 41.4% longer than the side). 40% is a significant amount and it will cause noticeable effects.

From my flimsy understanding of the SB ramp settings, I believe that the SB doesn't use ramps while doing a single circle command, and therefore the actual ramp value has nothing to do with this problem.

I was not implying that the SB cannot cut circles "smoothly" (a relative concept) - I was only trying to explain why the roughness would vary according to the position around the perimeter. And suggested speed changes to move away from resonances that may happen in the system. (By implication, the number of flutes on the cutter and the spindle speed could also be joining in this merry dance of motors and vibrations)

To get a feeling of your SB frame's natural frequencies, turn all motors off and tap the frame in various places (and directions) with a rubber hammer. Then feel the vibrations like you would do with a tuning fork. Any motor or cutter running near these natural frequencies will cause resonances. The high$$$ machines are much heavier and stiffer than the SB.

02-20-2003, 03:27 PM
thank you for the detailed explanation.
I see your point and now that I see it in diagrams, I understand about the motors changing speed.

The shopbot manual says "In addition to the ramps at the start and end of the circle, the speed of the motors change as the tool moves through the circle" (Look at that... it was there all the time!)

Maybe another thing to look at is the Threshold Angle. Chris and WingNut should look to see that theirs is set correctly.
This is the 4th parameter of the VT command. My manual states that the default is 115. This value is what determines when the tool will ramp or not. Angles larger or less sharp than 115 degrees should produce no ramping, while angles that are more sharp than 115 degrees will produce ramping.

If the Threshold Angle is set incorrectly this might cause ramping where it shouldn't be and add to the jerkiness. Just another shot in the dark...

02-20-2003, 05:51 PM
Thank you both for the insight to the dynamics of circle cutting. I will try all your suggestions to see if I can achieve the elusive perfectly flawless circle. I will report back after some more trial and error.

Many thanks

02-21-2003, 03:31 AM
One day, when I am really bored, I want to write a SB program that plays a recognisable tune.

I will send a bottle of red wine to anybody that gives me a SB file that plays the first two lines of our national anthem.